Method for detecting welding effect of lithium ion battery adapter plate

Abstract

The invention discloses a method for detecting the welding effect of an adapter piece of a lithium ion battery, and the method comprises the steps: setting test process parameters, and welding the adapter piece and a top cover according to the test process parameters; collecting a weld joint area, an ultrasonic signal and temperature information corresponding to the test process parameters, and establishing a theoretical penetration calculation model according to the test process parameters and the corresponding weld joint area, ultrasonic signal and temperature information; measuring the test fusion depth of a weld joint corresponding to the test process parameters, and inputting the test fusion depth into the theoretical fusion depth calculation model to obtain an actual fusion depth calculation model; and actual process parameters are input into the actual fusion depth calculation model to obtain the actual fusion depth, the actual fusion depth is compared with a standard value, and the welding effect is judged. According to the invention, the problems of resource waste and low detection efficiency caused by destructive detection adopted by the existing weld penetration detection are solved.

Description

Technical Field

[0001] This invention belongs to the field of lithium-ion battery technology, specifically relating to a method for detecting the welding effect of lithium-ion battery adapter pieces. Technical Background

[0002] Currently, the electrode tabs of square aluminum-cased batteries are typically welded together using ultrasonic welding to fuse multi-layer tab materials with metal adapter plates, followed by laser welding to bond the adapter plates to the terminals. However, during laser welding, welding abnormalities such as incomplete welds or failures to weld are prone to occur. Especially in cases of incomplete welds, although the weld mark on the surface may appear intact, the adapter plate and terminal plate are not truly fused, making the defect difficult to visually identify. Such incomplete weld defects can lead to increased internal resistance in the battery, and in severe cases, may even cause the battery to fail to charge properly or completely, directly affecting product yield and increasing production costs. Currently, the evaluation of welding effectiveness mainly relies on the detection of weld penetration depth. The traditional method for detecting weld penetration depth is metallographic analysis, which involves cutting and etching the welded part, and then observing the penetration depth of the weld area under a microscope. This method has the following problems: First, the detection process is time-consuming and labor-intensive; second, the detection method is destructive, which inevitably causes damage to the welded parts and waste of resources, and is not suitable for the needs of high-volume, high-efficiency welding quality inspection on mass production lines. Summary of the Invention

[0003] This invention provides a method for detecting the welding effect of lithium-ion battery adapter pieces, which solves the problem that existing methods for detecting the penetration depth of welded parts use destructive testing, resulting in resource waste and low testing efficiency.

[0004] This invention discloses a method for detecting the welding effect of a lithium-ion battery adapter piece, comprising: setting test process parameters, and welding the adapter piece to the top cover according to the test process parameters;

[0005] Collect the weld area, ultrasonic signal and temperature information corresponding to the test process parameters, and establish a theoretical penetration depth calculation model based on the test process parameters and the corresponding weld area, ultrasonic signal and temperature information;

[0006] The test penetration depth of the weld corresponding to the test process parameters is measured, and the test penetration depth is input into the theoretical penetration depth calculation model to obtain the actual penetration depth calculation model;

[0007] The actual process parameters are input into the actual penetration depth calculation model to obtain the actual penetration depth, and the actual penetration depth is compared with the standard value to determine the welding effect.

[0008] Furthermore, the step of setting test process parameters and welding the adapter plate to the top cover according to the test process parameters specifically involves:

[0009] The center power, ring power, welding speed, and defocusing amount of the welding process are set to perform laser welding on the same type of parts to be welded.

[0010] Furthermore, the acquisition of weld area, ultrasonic signal, and temperature information corresponding to the test process parameters includes:

[0011] The weld image corresponding to the test process parameters is acquired by the CCD detection module, the morphological information and pixel value of the image are obtained, and the area of ​​the weld is calculated.

[0012] The ultrasonic detection module collects the ultrasonic signal of the weld corresponding to the test process parameters, and determines whether the welding is successful based on the characteristic peak of the ultrasonic signal.

[0013] The infrared temperature detection module collects the temperature information of the welded parts corresponding to the test process parameters during the welding process. The temperature information of the welded parts includes the weld temperature and the surface temperature of the electrode post.

[0014] Further, the step of establishing a theoretical penetration depth calculation model based on the test process parameters and the corresponding weld area, the ultrasonic signal, and the temperature information includes:

[0015] Based on the center power, ring power, welding speed, and weld area during the welding test, calculate the line energy density and surface energy density during the welding process.

[0016] Based on the linear energy density, the surface energy density, and the temperature difference between the weld and the pole surface, a multivariate nonlinear regression model is used for fitting to obtain the theoretical penetration depth calculation model.

[0017] Further, the step of measuring the test penetration depth of the weld corresponding to the test process parameters, inputting the test penetration depth into the theoretical penetration depth calculation model, and obtaining the actual penetration depth calculation model includes:

[0018] The actual weld penetration depth under the test process parameters is detected by metallographic examination, and the tested penetration depth is input into the theoretical penetration depth calculation model for mathematical fitting to obtain the coefficients to be fitted and the random error.

[0019] Furthermore, after inputting the tested melt depth into the theoretical melt depth calculation model and obtaining the actual melt depth calculation model, and before inputting the actual process parameters into the actual melt depth calculation model to obtain the actual melt depth, the method further includes:

[0020] The accuracy of the actual melting depth calculation model was verified and optimized.

[0021] Furthermore, the accuracy of the actual melting depth calculation model is verified and optimized, including: calculating the determination coefficient of the actual melting depth calculation model; if the determination coefficient is greater than 85, then the actual melting depth calculation model has a good fitting effect.

[0022] If the residuals are randomly distributed when performing residual analysis on the actual melting depth calculation model, then the actual melting depth model has no systematic bias.

[0023] Furthermore, before inputting the actual process parameters into the actual melt depth calculation model to obtain the actual melt depth, the process further includes:

[0024] The weld seam image corresponding to the actual process parameters is acquired by the CCD detection module, the morphological information and pixel value of the image are obtained, and the area of ​​the weld seam is calculated.

[0025] The ultrasonic detection module collects the ultrasonic signal of the weld corresponding to the actual process parameters, and determines whether the welding is successful based on the characteristic peak of the ultrasonic signal.

[0026] The infrared temperature detection module collects the temperature information of the welded parts corresponding to the actual process parameters during the welding process. The temperature information of the welded parts includes the weld temperature and the surface temperature of the electrode post.

[0027] Furthermore, the step of inputting actual process parameters into the actual melt depth calculation model to obtain the actual melt depth includes:

[0028] The actual process parameters during the actual welding process are measured, including: center power, ring power, welding speed, and decoking amount;

[0029] Based on the actual process parameters and the area of ​​the weld, calculate the linear energy density and surface energy density during the welding process;

[0030] The actual melting depth is obtained by inputting the linear energy density, the surface energy density, and the temperature difference between the weld and the pole surface into the actual melting depth calculation model.

[0031] Furthermore, the step of comparing the actual penetration depth with a standard value to determine the welding effect includes:

[0032] If the actual penetration depth is within the range of 600±100μm, the welding effect is good; if it is not within the range of 600±100μm, the welded part should be rejected for verification.

[0033] The beneficial effects of this invention are as follows: By collecting weld area, ultrasonic signals, and temperature information generated during the welding process, and combining this with the measured penetration depth to establish an actual penetration depth calculation model, the actual penetration depth of the weld can be predicted without damaging the welded parts. This achieves non-destructive testing of welding quality, reduces testing costs and resource consumption, and is suitable for welding quality monitoring in mass production. Based on the established actual penetration depth calculation model, the predicted penetration depth can be obtained by substituting the actual process parameters into the model. This prediction can then be compared with the standard value, facilitating rapid judgment of the welding effect, reducing testing time, and improving the testing efficiency of the production line and the real-time performance and accuracy of yield control. Attached Figure Description

[0034] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 A flowchart illustrating a method for detecting the welding effect of a lithium-ion battery adapter plate according to an embodiment of the present invention;

[0036] Figure 2 A curve fitting the predicted melt depth to the actual melt depth;

[0037] Figure 3 This is a residual graph showing the difference between the measured and predicted melt depths.

[0038] Figure 4 This is another flowchart illustrating a method for detecting the welding effect of a lithium-ion battery adapter piece, provided in an embodiment of the present invention. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] This application provides a method for detecting the welding effect of lithium-ion battery adapter pieces, such as... Figure 1 As shown, the detection methods include:

[0041] S1 sets the test process parameters and welds the adapter plate to the top cover according to the test process parameters;

[0042] S2 collects the weld area, ultrasonic signal, and temperature information corresponding to the test process parameters;

[0043] S3 establishes a theoretical penetration depth calculation model based on the test process parameters and the corresponding weld area, ultrasonic signal and temperature information;

[0044] S4 measures the test penetration depth of the weld corresponding to the test process parameters, inputs the test penetration depth into the theoretical penetration depth calculation model, and obtains the actual penetration depth calculation model;

[0045] S5 inputs the actual process parameters into the actual penetration depth calculation model to obtain the actual penetration depth, and compares the actual penetration depth with the standard value to determine the welding effect;

[0046] The S1 process parameters include setting the welding process center power: 3.0±1KW, ring power: 1.5±1KW, welding speed: 200mm / s-800mm / s, and defocusing amount: 1±3mm, and performing laser welding on the same type of parts to be welded.

[0047] S2 collects the weld area, ultrasonic signal, and temperature information corresponding to the test process parameters, including: acquiring weld images corresponding to the test process parameters through a CCD detection module, obtaining the image morphology information and pixel values, and calculating the weld area (S / mm²). 2 );

[0048] The ultrasonic detection module collects the ultrasonic signal of the weld corresponding to the test process parameters, and determines whether the welding is successful based on the characteristic peak of the ultrasonic signal.

[0049] By collecting ultrasonic signals from the weld and analyzing the amplitude of the ultrasonic signals, it can be seen that when the weld is successful, the amplitude of the main reflection peak of the echo signal is high and has good consistency; when the weld is incomplete or not fused, due to acoustic impedance mismatch, the amplitude of the main reflection peak is significantly reduced or distorted, or abnormal multiple secondary reflections occur.

[0050] The temperature information of the welded parts corresponding to the test process parameters during the welding process is collected by an infrared temperature detection module. The temperature information of the welded parts includes the weld temperature (T). w ( / ℃) and pole surface temperature T c / ℃.

[0051] S3 establishes a theoretical penetration depth calculation model based on test process parameters and corresponding weld area, ultrasonic signals, and temperature information; including:

[0052] According to the input parameters (welding center power (P) c / KW), welding ring power (P) r / KW), welding speed (V / mm / s), defocusing amount (d / mm), welding area (S / mm) 2 ), weld temperature (T) w / ℃), electrode surface temperature (T) c The linear energy density and surface energy density are calculated by ( / ℃), where the linear energy density is the input energy per unit length and the surface energy density is the input energy per unit area.

[0053] Total power: P = P c +P r Unit: KW

[0054] Linear energy density:

[0055] Surface energy density: Based on the defocusing amount d, and the temperature difference ΔT = T between the weld temperature and the electrode surface temperature. w -T c A preliminary fitting of the relevant process parameters using a multivariate nonlinear regression model yields the theoretical melt depth calculation model:

[0056]

[0057] in:

[0058] h represents the predicted weld penetration depth;

[0059] β0, β1, β2, β3, and β4 are the coefficients to be fitted, which are determined through regression analysis of experimental data.

[0060] ΔT is the temperature difference between the weld and the surface temperature of the pole.

[0061] This is a random error, which is used to correct the error between the actual melting depth and the theoretical melting depth.

[0062] S5 measures the test penetration depth of the weld corresponding to the test process parameters, inputs the test penetration depth into the theoretical penetration depth calculation model, and obtains the actual penetration depth calculation model, including:

[0063] Metallographic examination was performed on the weld under the test process parameters to obtain the test penetration depth. The test penetration depth and the corresponding test process parameters were then substituted into the theoretical penetration depth calculation model to obtain the fitting coefficients and random error values. The specific experimental data are shown in the table below:

[0064]

[0065]

[0066] Based on the above data, regression calculations were performed to determine the fitting coefficients and random error values, ultimately yielding the following actual melting depth calculation model.

[0067]

[0068] Compared to traditional weld penetration testing, metallographic testing of samples, by collecting weld area, ultrasonic signals and temperature information generated during the welding process, and combining this with the measured weld penetration to establish a weld penetration calculation model, can predict the actual weld penetration without damaging the welded sample, thus achieving non-destructive testing of weld quality.

[0069] After inputting the test melt depth into the theoretical melt depth calculation model in S5 and obtaining the actual melt depth calculation model, the detection method further includes:

[0070] The accuracy of the actual melting depth calculation model was verified and optimized.

[0071] The accuracy of the regression model is verified and optimized by calculating the coefficient of determination (R-Sq) and performing residual analysis. A coefficient of determination (R-Sq) greater than 85 indicates that the actual melt depth calculation model has a good fit and can be used to predict melt depth. Residual analysis of the actual melt depth model shows that if the residuals are randomly distributed, it proves that the actual melt depth model has no systematic bias.

[0072] By setting a criterion for the coefficient of determination R-Sq (e.g., R-Sq > 0.85), the model's explanatory power for variations in sample data is evaluated, ensuring that the established actual weld penetration model has high predictive accuracy under current process conditions and is suitable for welding quality assessment scenarios. Residual analysis is used to determine the distribution of model errors. When the residuals are randomly distributed and show no obvious trend, it indicates that the model is not affected by systematic biases, improving the model's stability and reliability under different samples and process batches. This method can identify the risk of welding misjudgment caused by model errors, ensuring the consistency and repeatability of prediction results, thereby reducing false weld misjudgments and improving welding inspection yield.

[0073] Figure 2 The fitted curve between the predicted and actual weld depths shows that most points are distributed near the fitted line with no significant deviation, indicating a good fit without systematic bias, and can be used to predict weld depth. This demonstrates that the regression model can explain 90.4% of the actual weld depth variance, showing a high degree of fit. The fitted curve can be used as a predictive model in actual processes, allowing for the assessment of whether the welding effect meets standards based on the predicted values.

[0074] Figure 3The residual plots for measured and predicted melt depths show that the points in the normal probability plot are distributed primarily along the diagonal, indicating that the residuals approximately follow a normal distribution, satisfying the basic assumption of the regression model regarding the normality of the error term. In the residual fit value plot, the residuals are generally randomly distributed without a significant nonlinear trend, indicating that the model's linear assumption is largely valid. In the residual histogram, the residuals show a skewed distribution, with most residuals concentrated in the -20 to +30 range, indicating that the prediction error is mainly within a controllable range. Although not perfectly symmetrical, the shape is relatively smooth, indicating good model stability. The residual observation sequence plot shows a high-frequency fluctuation pattern with no systematic upward or downward trend, indicating that the residuals are random and there is no time-series correlation or systematic model bias; overall, it conforms to the regression residual independence assumption.

[0075] Figure 4 This is another flowchart illustrating a method for detecting the welding effect of a lithium-ion battery adapter plate according to an embodiment of the present invention, as shown below. Figure 3 As shown, S5 inputs the actual process parameters into the actual melt depth calculation model to obtain the actual melt depth, including:

[0076] S51 performs laser welding according to actual process parameters;

[0077] S52 acquires weld images corresponding to the actual process parameters through the CCD detection module, obtains the morphological information and pixel values ​​of the images, and calculates the area of ​​the weld.

[0078] S53 acquires the ultrasonic signal of the weld corresponding to the actual process parameters through the ultrasonic detection module, and determines whether the welding is successful based on the characteristic peak of the ultrasonic signal.

[0079] S54 collects the temperature information of the welded part corresponding to the actual process parameters during the welding process through the infrared temperature detection module. The temperature information of the welded part includes the weld temperature and the surface temperature of the pole.

[0080] S55 calculates the linear energy density and surface energy density during the welding process based on the actual process parameters and the area of ​​the weld; and inputs the linear energy density, the surface energy density, and the temperature difference between the weld and the pole surface into the actual penetration depth calculation model to obtain the actual penetration depth.

[0081] Meanwhile, by establishing a periodic data update mechanism, key data of the welding process (including actual process parameters, ultrasonic signals, temperature information, weld area, and some manual sampling data of actual weld depth) are continuously collected during the production process. The newly collected data is incorporated into the model training or retraining process, and the original model is adaptively adjusted through refitting, incremental learning, or model transfer, so that the model can dynamically reflect the real physical process under the current production conditions and avoid fitting errors caused by equipment aging, material fluctuations, etc.

[0082] The actual penetration depth is compared with a standard value to determine the welding effect. This includes: if the actual penetration depth is within 600±100μm, the welding effect is considered good; if it is not within this range, the welded part is rejected for further verification. Based on the established actual penetration depth calculation model, the predicted penetration depth can be obtained by substituting the actual process parameters into the model. This predicted penetration depth is then compared with the standard value, allowing for rapid assessment of the welding effect. This reduces inspection time and improves the inspection efficiency of the production line, as well as the real-time performance and accuracy of yield control.

[0083] The embodiments described above are merely examples of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the protection scope of this application.

Claims

1. A method for detecting the welding effect of lithium-ion battery adapter pieces, characterized in that, include: Set the test process parameters, and weld the adapter plate and the top cover according to the test process parameters; Collect the weld area, ultrasonic signal and temperature information corresponding to the test process parameters, and establish a theoretical penetration depth calculation model based on the test process parameters and the corresponding weld area, ultrasonic signal and temperature information; The test penetration depth of the weld corresponding to the test process parameters is measured, and the test penetration depth is input into the theoretical penetration depth calculation model to obtain the actual penetration depth calculation model; The actual process parameters are input into the actual penetration depth calculation model to obtain the actual penetration depth, and the actual penetration depth is compared with the standard value to determine the welding effect.

2. The method for detecting the welding effect of a lithium-ion battery adapter plate as described in claim 1, characterized in that, The process of setting test process parameters and welding the adapter plate to the top cover according to the test process parameters is as follows: The center power, ring power, welding speed, and defocusing amount of the welding process are set to perform laser welding on the same type of parts to be welded.

3. The method for detecting the welding effect of a lithium-ion battery adapter plate as described in claim 2, characterized in that, The acquisition of weld area, ultrasonic signal, and temperature information corresponding to the test process parameters includes: The weld image corresponding to the test process parameters is acquired by the CCD detection module, the morphological information and pixel value of the image are obtained, and the area of ​​the weld is calculated. The ultrasonic detection module collects the ultrasonic signal of the weld corresponding to the test process parameters, and determines whether the welding is successful based on the characteristic peak of the ultrasonic signal. The infrared temperature detection module collects the temperature information of the welded parts corresponding to the test process parameters during the welding process. The temperature information of the welded parts includes the weld temperature and the surface temperature of the electrode post.

4. The method for detecting the welding effect of a lithium-ion battery adapter piece as described in claim 3, characterized in that, The step of establishing a theoretical penetration depth calculation model based on the test process parameters and the corresponding weld area, ultrasonic signal, and temperature information includes: Based on the center power, ring power, welding speed, and weld area during the welding test, calculate the line energy density and surface energy density during the welding process. Based on the linear energy density, the surface energy density, and the temperature difference between the weld and the pole surface, a multivariate nonlinear regression model is used for fitting to obtain the theoretical penetration depth calculation model.

5. The method for detecting the welding effect of a lithium-ion battery adapter piece as described in claim 4, characterized in that, The process of measuring the test penetration depth of the weld corresponding to the test process parameters, inputting the test penetration depth into the theoretical penetration depth calculation model, and obtaining the actual penetration depth calculation model includes: The actual weld penetration depth under the test process parameters is detected by metallographic examination, and the tested penetration depth is input into the theoretical penetration depth calculation model for mathematical fitting to obtain the coefficients to be fitted and the random error.

6. The method for detecting the welding effect of a lithium-ion battery adapter piece as described in claim 1, characterized in that, After inputting the tested melt depth into the theoretical melt depth calculation model and obtaining the actual melt depth calculation model, and before inputting the actual process parameters into the actual melt depth calculation model to obtain the actual melt depth, the process further includes: The accuracy of the actual melting depth calculation model was verified and optimized.

7. The method for detecting the welding effect of a lithium-ion battery adapter piece as described in claim 6, characterized in that, The accuracy verification and optimization of the actual melt depth calculation model include: Calculate the determination coefficient of the actual melting depth calculation model. If the determination coefficient is greater than 85, the actual melting depth calculation model has a good fitting effect. If the residuals are randomly distributed when performing residual analysis on the actual melting depth calculation model, then the actual melting depth model has no systematic bias.

8. The method for detecting the welding effect of a lithium-ion battery adapter plate as described in claim 1, characterized in that, Before inputting the actual process parameters into the actual melt depth calculation model to obtain the actual melt depth, the method further includes: The weld seam image corresponding to the actual process parameters is acquired by the CCD detection module, the morphological information and pixel value of the image are obtained, and the area of ​​the weld seam is calculated. The ultrasonic detection module collects the ultrasonic signal of the weld corresponding to the actual process parameters, and determines whether the welding is successful based on the characteristic peak of the ultrasonic signal. The infrared temperature detection module collects the temperature information of the welded parts corresponding to the actual process parameters during the welding process. The temperature information of the welded parts includes the weld temperature and the surface temperature of the electrode post.

9. The method for detecting the welding effect of a lithium-ion battery adapter piece as described in claim 8, characterized in that, The step of inputting actual process parameters into the actual melt depth calculation model to obtain the actual melt depth includes: The actual process parameters during the actual welding process are measured, including: center power, ring power, welding speed, and decoking amount; Based on the actual process parameters and the area of ​​the weld, the linear energy density and surface energy density during the welding process are calculated; the linear energy density, the surface energy density, and the temperature difference between the weld and the pole surface are input into the actual penetration depth calculation model to obtain the actual penetration depth.

10. The method for detecting the welding effect of a lithium-ion battery adapter piece as described in claim 9, characterized in that, The step of comparing the actual penetration depth with the standard value to determine the welding effect includes: If the actual penetration depth is within the range of 600±100μm, the welding effect is good; if it is not within the range of 600±100μm, the welded part should be rejected for verification.

Images